Mr. Chairman and Gentlemen:—The brief time I shall ask your attention this evening precludes the possibility of my devoting any attention to all of the various instruments and appliances for indicating and recording fluid pressure. I shall therefore limit myself to the consideration of that class of instruments which have been brought forth as reliable, and applicable to pressures varying from an absolute zero to that of a few thousand pounds per square inch (they forming those of the most extended manufacture and use in the daily arts), and conclude with a few words on their application and desirability.
Instruments for only indicating pressure have been made under two general types—that of the bent tube, after Bourdon, and that of the vibrating diaphragm, from the German. These move according to the fluctuating pressure applied, and such movement I term “original travel." The pressure upon them is determined by the mercury column as a standard, and their scales and dials are laid off accordingly. Starting with the assumption that the mercury column is accurately constructed and properly used, we have to consider the form and appropriateness of the metals employed in these two types for sustaining without change or "set" constantly varying strains and shapes so that similar exposure to pressure shall result in similar travel and dial reading. The two types mentioned have, as a rule, predominated for indicating purposes in about equal numbers, and been accepted as equally trustworthy and reliable by the masses, who neglect to question or probe beneath the surface. When we consider that the duty is that of a spring, based upon the resilience of the metal composing the same, it seems somewhat strange that the Bourdon tube, made of brass composition, should even be expected to act equally with the properly tempered disks of steel. Ordinarily, where the vicissitudes of its use admit of the same, we invariably employ tempered steel when we want a spring, as we know more about making it of that metal than we do of alloys of copper, tin, etc.; and in practice the failure in the Bourdon tube is due to the yielding of the metal composing it, while the steel diaphragm, when well made, has an endurance which nothing but tempered steel can give. Unfortunately, however, these diaphragm instruments have been poorly made both as to the spring and the multiplying mechanism for throwing the hand around the dial. Nor does the most difficult portion of the problem consist in simply providing a reliable spring, because it is equally important to have a uniform and lively “travel " from the pressure itself, in order to dispense to the utmost degree with the introduction of multiplying gears, which multiply errors as readily as they do the "original travel." Of equal importance is it, also, that persistence of the “travel" shall be able to overcome the friction to be encountered in moving the multiplying and indicating mechanism. This persistency of change, due to the variable pressure in the tube, is extremely slight, and in the diaphragm very great. In the latter it is determined by multiplying the exposed area of the diaphragm by the pressure in pounds per square inch it is sustaining. As friction must be overcome and work done without disturbing the accuracy of the instrument in its movement, this "persistency of travel” possessed by the diaphragm alone is a very important feature. Therefore it is sufficient for our present purposes that we take into account only the consideration of steel diaphragms and their superincumbent mechanism for giving multiplied travel to the hand upon the dial.
By custom, for commercial purposes, they have been made of French and English steel, varying in thickness according to the pressure to be sustained, and cut into small disks two and one- half inches in diameter, and corrugated slightly to increase their displacement under various pressures. Such disks give a travel of only one thirty-second of an inch. No reason is to be found why this small diameter was selected, or why it has always been adhered to. The total "original travel" from the pressure itself must be proportionate to the diameter, and while a greater travel is always ultimately required for dial purposes, it has only been obtained through high multiplication, by gearing of the small “original travel," and at the expense of durability and accuracy.
My attention was called to this matter in 1869 or 1870, and soon discovering that the diaphragms in use were improperly constructed, I at once introduced two improvements, which consisted in corrugating the diaphragms convexly, and increasing the dimensions and amplitude of such corrugations from the centre to the circumference. This resulted in more than doubling the "original travel” previously obtainable from a given diameter and thickness of disk under similar pressure. Thus encouraged, I increased the diameter from two and one-half to six and one-half inches, obtaining an "original travel" directly from the pressure itself, without multiplication, of from three eighths to one-half an inch. With an abundance of travel, I also had a "persistency of travel," unaffected by any reasonable amount of friction to be encountered in moving an indicating or a recording mechanism. In fact, at this point started the present recording apparatus, after the continuous failure of many others, here and abroad, in their attempt to get a competent and durable source of “original travel" suitable for a pressure-measuring and recording mechanism. That the problem then received no meager solution is attested by the fact that the type of pressure recorder here exhibited has been in extensive use since 1S75. On the other hand, while the pressure recording instrument has been steadily making a record of reliability for itself, the simple indicating gauge has as steadily pursued a retrograde course, until now the commercial instrument is so carelessly standardized, if standardized at all, that reliance upon it becomes positively dangerous, and cannot be too strongly deprecated. In fact, the eminent engineer, the late J. C. Hoadley, as one of a committee appointed to compile a code of rules to be adopted in making steam-boiler tests, specifies an Edson recording gauge and its record as a good check upon the indications of the ordinary steam gauge. So in the report for August, 1885, of the inspectors of the Hartford Steam Boiler Insurance Company, they give: 255 pressure-indicating gauges defective—31 dangerously so—and 2 boilers without them; report for April of same year gives i6g defective, 42 dangerously so.
The instrument referred to, briefly described, consists of a metal base A, enclosing beneath it a tempered diaphragm C, so arranged that when the fluid enters the space D between the spring and the cap E, forming the chamber, the spring is deflected upwardly. The recording apparatus is mounted on the top of the base A, and the movement of this diaphragm C is transmitted through the arms Hi and Hi on the rock shaft H, by means of the connecting bar G, to the vertical moving pencil carrier in front, thrown thereby about six times the original travel of the diaphragm. Simultaneously therewith,
motion from the same rock shaft H moves the hand Mi before the dial M. A special clock mechanism revolves the receiving reel Ki, contributing the element of time to the chart drawn from the reservoir reel A' on the left, beneath the recording pencil. Ordinarily this reservoir reel K contains a supply of charts for thirty days. A glass dome surmounting the whole enables inspection and excludes dust, moisture, etc. The supplemental, adjustable arm O upon the rock shaft H acts both as a circuit closer for an electro-magnetic alarm, and operates a mechanical alarm usually provided with 'each instrument
This form has been found successful in recording pressure from two pounds per square inch for blast furnaces, to (1200) twelve hundred pounds per square inch used in pumping oil. For recording temperatures of drying-rooms, etc., a supplemental diaphragm is employed to increase the travel, owing to the low coefficient of expansion possessed by the fluids used for the various temperature ranges.
I take it that, of all the various uses to which pressure-recording instruments are put, none equals in importance, practically, that of being called upon to stand sentinel over the pressure of steam. Certainly no other explosive element is encountered so frequently, or enters so largely into our domestic economy; and equally true is it that no other is produced in such ignorance or left to such carelessness. Attendant safety and economy are almost entirely left to become matters of chance, and while the average steam user readily accords due homage to a pound of gunpowder known to be stored about his establishment, willingly assuming it to be dry, he cannot be made to believe that each cubic foot of water in his boiler under sixty pounds of steam contains the explosive energy of that same pound of gunpowder. Equally true is it from an economical point of view that the Stoker who will labor the cheapest is the cheapest laborer to his mind; and while ready to denounce as extortion a demand for an increase of wages of even five per cent, on the part of the boiler attendant, he readily pays, unquestioned, any coal bill at sight, though its excess may more than equal the wages of the man.
A well-constructed boiler, proper firing, moderate draught and steady feed—these form the basis of the best economy. The pressure of steam in the boiler will vary from two causes—upwardly from excess of supply over demand, and downwardly from the demand exceeding the supply. Now, if the demand be constant, there is no reason why the supply should not be constant also. If, on the other hand, the demand for steam is variable, does it follow that the supply should also become inconstant? Inequality in demand may be regular and right, but inequality in supply is never right, for the conditions which give rise to it are almost invariably attributable to carelessness or ignorance or inexperience. This is, to a large extent, the natural consequence of the peculiar conditions under which steam is generated. The idea prevails that, to obtain steam, coal must be burned, and in order to practise frugality the cheapest labor must handle it in the fire-room; while by the same intelligence the conclusion is reached that the engine must be presided over, for economy’s sake, by a skillful and experienced man—of necessity! But we are not all ignorant of the relative stewardship of the boiler with the coal supplied it, and the engine with the steam it gets; and the time may not be far distant when, copying the practice from the Caspian Sea and substituting oil for coal, we may not only better use our fuel, but bring mechanical appliances in play which will be much more reliable than careless or ignorant human agents can be. The most ordinary intellect would at once intuitively, as it were, recognize the necessity for supplying the stream of oil in just the proper quantity uninterruptedly, and would strive to preserve the proper adjustments of the apparatus for its accomplishment; while the same person, if using coal as a fuel, would entirely ignore the extra necessity for vigilance and attention required to maintain a steady supply of it, and a uniform rate of its distillation into the gases of combustion. That from five to thirty per cent, of the fuel used for a given steam production can be saved or wasted, never occurs to the average fireman or his employer; and, therefore, the fidelity to duty imposed upon him by the knowledge of the operation of the pressure-recording gauge compels him to give the utmost attention to his work, and so, incidentally, to practise economy.
Once, during a three years' cruise in the "service," just after the close of the war, I took the trouble to compare the amounts of coal burned by the different watches for a continuous period of ninety days' steaming. There were three watches of firemen and four of engineers. Taking the coal consumed per knot by the most economical officer as a standard of comparison with the others, the results gave a difference, respectively, of eight, twenty-two and thirty-odd per cent. That is, thirty per cent, more coal was consumed per knot made when one engineer was on duty than was consumed with another.
I mention the matter not as a specimen of poor engineering, but for the purpose of showing it to be possible to get rid of an excess of thirty per cent, of coal without generating the heat and steam due to its consumption. Seldom does it occur to the proprietor that his steam boiler has a direct passage from the furnace door to the smoke-stack, through which the heat from the furnace may pass and become fugitive without his having properly arrested its motion, or taken from it that to which he is entitled. Nor does he realize or know how little of the theoretical value of the coal it is possible for him to obtain under the best of conditions and with the most modern pattern of boiler skillfully fired. Ignorant of how little he could get under such circumstances, he little realizes how much he loses under the average practice. Still he continues to look for cheaper labor, and does not complain of larger coal bills.
The extravagant use of fuel is but one of the sources of loss occasioned by carelessness and ignorance, to be remedied, to a great extent, by the simple adoption of that form of pressure gauge that shall make a time record of its indications. That unobserved loss which comes gradually but with unerring certainty in the form of the impairment of the boiler's condition is, perhaps, of paramount importance to mere loss of fuel. One affects financial results solely, and the other deals directly with safety of life and property as well. Therefore, whatever unnecessarily injures or weakens a boiler should be avoided voluntarily, or, by legal enactment, prevented, as no one has the right to hazard either life or property; and the employment, wherever fluids are maintained under great compression, of an instrument which shall record upon paper what that compression is and has been, is not only scientifically expedient and useful, but becomes a manifest duty towards those who are in any way endangered thereby.
Since the discovery of the value of steam for a motive power, and for various other uses rendering steam boilers a necessity, no means have heretofore been produced by which reliable records of the pressure to which such boilers were from time to time subjected could be obtained automatically, and to the very absence of which, in a great measure, are to be attributed the constantly recurring boiler explosions, with the usual verdict of" nobody to blame" when investigations are held. Without these records of the pressure carried, the fireman, who is -frequently left in charge night as well as day, may take as much risk in carrying steam as suits his convenience or ignorance, regardless of economy and safety, and without fear of discovery. Before closing I wish to show you some stereopticon views, and the accompanying records of boiler pressure carried in the places severally named, and automatically written by means of Edson's time and pressure-recording steam gauges, are half-size
reproductions from the originals.
The late eminent Chief of the Bureau of Steam Engineering, W. W. W. Wood, wrote as follows, in November, 1870: “I attach great importance to your recording gauges, by which extravagance in use of fuel may be positively detected, and accidents, which too frequently involve the loss of life and property, may be avoided." It hardly seems as though an intelligent man could say otherwise; but still they do exist.
In conclusion, does it not seem that our age is too far advanced in scientific improvements longer to permit boiler explosions and consequent loss of human life through negligence, ignorance and sordid notions of economy? I am warranted, by the indisputable records of charts for years, in saying that the Edson pressure-recording gauge writes a constant and unimpeachable record of the facts of variations of fluid pressure, and without which no steam boiler “plant” can be conducted either intelligently, economically, or safely; that there is no longer excuse for mistakes or ignorance resulting in loss of life or property either on land or water, for, by the use of this pressure-recording gauge, the utmost immunity from danger, greatest economy in fuel, and the fullest information are obtainable. Surely such instruments are not only cheap in the end, but, where lives are in jeopardy, they are imperative and of moral obligation.
DISCUSSION.
Passed-Assistant Engineer W. F. Worthington, U. S. N.—Mr. Chairman and Gentlemen:—No one who has had charge of a marine engine will deny that a record of its performance is of great value. The first thing an engineer wants to know when coming on watch is how the machinery has been working, not only with regard to cool journals and smoothness of movement, but every detail of pressure, speed, amount of coal, etc., in order to know what requires his first care, and what his principal duty will be for the next four hours. If the record of the log as now kept—with only approximate accuracy—is of real value, it is evident that greater accuracy would increase its value. As long as the officer on watch can remain on the engine-room platform, and the steam pressure does not vary much, he can judge pretty well the average pressure for the hour; but let his attention be diverted for fifteen or twenty minutes, or longer, by a hot journal or a refractory feed pump, and the record of steam for that hour will not be worth much for scientific purposes. If the revolutions which are recorded automatically are the same, and also the vacuum, etc., he will probably assume that the average steam pressure was also the same; but this is at best a guess, and might be a wrong one, owing to a change in the direction or force of the wind which would allow the same number of revolutions to be made with less pressure. One of the most important uses of the recording gauge is to correct the error of observation due to the observer. I believe this error to be as much as two or three pounds. If this is cumulative—that is, if several observers come on watch in succession, each with a tendency to mark the pressure a little higher than it is—the steam may have been going down very gradually from neglected fires or change in quality of coal without the fact becoming apparent, until finally the speed of the ship is materially affected, and several hours are required to bring things back to their proper condition. The general tendency of the pressure, whether up or down, cannot be observed in so short a time as four hours. At the end of that time another observer comes on duty, and it is not known whether he records too high or too low, so that eight hours pass without affording a clue to the desired information; and this is sufficient time for the fires to get dirty if the coal is bad, or for the water-tender to run the water too high in all the boilers, so as to have time for a quiet smoke on the morning watch.
One of the first results of the introduction of recording gauges on shipboard would be more uniform firing, with all the well-known advantages to be derived there from. A diagram, such as is drawn by the Edson gauge, appeals to the intelligence of the most untutored fireman, and there would be a friendly rivalry between the different watches to produce the best results, just as there now is to get the greatest number of revolutions recorded by the counter. Another advantage would be, that if one watch were weaker than another, the fact would become apparent by examining the diagrams for several days, and the men could then be rearranged to produce more even firing. Of course this can be done now, but not so well, owing to the doubt of the accuracy of the steam pressures recorded by different observers. Another advantage would be the greater accuracy in determining the mean I. H. P. for the day.
Up to the present time there is no reliable continuous steam-engine indicator, and it is necessary to average the steam pressure for twenty-four hours and then get an indicator diagram with the steam pressure, cut-off, etc., the same as the average shown on the log for the twenty-four hours. If the pressures recorded for each hour are not exact to several pounds, it may easily happen that the errors do not balance, but all lie on one side of the truth, and the mean I. H. P. would be deceptive.
Every one who has attempted to draw conclusions from the data recorded in the steam log, has felt the inconvenience which arises from the uncertainty with regard to the steam pressure. The vacuum, the revolutions, temperature of the feed, weight of coal, etc., can be ascertained with considerable accuracy or within fixed and comparatively narrow limits; but the steam fluctuates widely—as much as ten pounds when the working pressure is sixty—and therefore the conclusions drawn can only be relied upon as correct within these wide limits. It has frequently been observed that duplicate engines, the same in every detail, apparently, produce different I. H. P. A recording gauge, applied to each in turn, would soon determine if this difference was real, and thus advance us one good step towards the solution of an interesting and important problem. The engines of warships should be of the highest attainable efficiency, and, in order to make them so, it is necessary to know their defects. The best and, indeed, only way to discover these defects is to get accurate data of the working under every circumstance, especially at sea; and the way to get this data is to have it recorded automatically, because no one can devote his whole time to watching the various instruments. The only objection I ever heard to the use of a recording gauge was that it was unnecessary, because the officer on watch was trustworthy and it would not be right to set a machine to watch him. A complete answer to this objection is that one who does his work well does not care who knows it. One in charge of an engine has many other important duties besides watching the steam gauge, and the recording and alarm apparatus would materially assist him. The attachment for recording the number of revolutions simultaneously with the pressure is a valuable addition to the Edson gauge. The pressure and revolutions often vary independently of each other, and when the latter vary—a fact at once detected by a practical ear before any of the other instruments note it—then the cause could immediately be inquired into and the remedy applied. Although, in time of peace, every run made by a war vessel should be a trial trip in one sense, the value of the recording gauge would appear highest when a full-power trial is made. On such occasions the danger of hot journals, breakages, etc., is greatest, and every one on duty is fully occupied in causing the machinery to do its utmost, so that just when there is most need of getting correct data there is least time to collect it.
Lieutenant A. M. Knight, U. S. N.—Mi: Chairman and Gentlemen:—I have been requested, in connection with the subject of the evening, to give a description of the pressure-recording instruments used in the experimental firing of great guns.
The difficulty in measuring the pressure developed by a powder charge burning in the bore of a gun arises from: 1st. The magnitude of the pressure involved; 2d. The suddenness with which it is developed; 3d. The brevity of the time through which it acts; 4th. The rapidity with which it varies during that time, under the influence of conditions of temperature, motion, etc., of which we have but a very limited knowledge.
The magnitude of the pressure involved, which commonly varies from twelve to twenty tons to the square inch, and in exceptional cases rises to twenty-five and even thirty tons, would of itself put out of the question the employment of any ordinary gauge. Various instruments have been invented which meet this and the other sources of difficulty enumerated above with more or less success. Of these, those which are in practical use at the present time may be divided into two general classes: 1st. Those in which the pressure is measured by opposing to it a known high resistance—usually the resistance of a ductile metal like lead or copper to deformation by cutting or crushing—and 2d. Those in which the pressure is measured by its effect in imparting velocity to a body of known dimensions and mass.
Of the first class, the most important examples are the Rodman, the Woodbridge and the crusher gauges, and the Deprez manometric balance. To the second class belong the Noble chronoscope, the Sebert velocimetre and the Sebert self-registering projectile.
The Rodman gauge (Fig. 1) was invented about 1857 by Colonel Rodman, of the United States Army, and embodies the general principles of the more recent Woodbridge and crusher gauges.
It consists of a steel cylinder, A, which screws into the wall of the gun or into a plug inserted in the powder chamber, its head being flush, or nearly flush, with the surface of the bore. Within the cylinder is fitted a piston, B, one end of which, when the gauge is in place, is just inside the head of the gauge and free to be acted upon by the gases developed in the bore on firing. To the other end of the piston is attached a knife-edge, E, of tempered steel, and in light contact with this is placed a disc of copper, C, resting upon the screw plug which closes the base of the cylinder.
A copper cup, D, placed over the head of the piston, acts as a check to prevent the gases from entering the gauge.
When a pressure is developed in the bore of the gun, the piston of the gauge is forced inward, and the knife-edge enters the copper disc to a depth dependent upon the amount of the pressure. The gauge is then removed from the gun, the disc taken out, and the dimensions of the cut measured and compared with those of cuts made upon the same or similar discs by the same knife-edge and piston when subjected to known pressure in a testing machine. It is assumed that equal pressure in the machine and in the gun will produce equal effects upon the copper discs. As will be explained later, this assumption involves an error.
The Woodbridge gauge differs from the Rodman in the cutter employed. The knife-edge of the Rodman system is replaced by a narrow groove, with a cutting edge, running spirally around a concave conical surface on the end of the piston. The outer rim of this surface is in contact with the circumference of the copper disc when the gauge is ready for use. As the piston is forced in upon the disc by the pressure of discharge, the disc is compressed from its circumference towards its centre, the amount of compression being indicated by the length of the spiral cutting edge, which leaves its mark upon the disc. This record is compared with that made upon similar discs by the same piston acting under known forces.
In the crusher gauge, the cutters of the two preceding systems are discarded. The piston head is plain, and between it and the base plug of the gauge is placed a copper cylinder whose length is accurately measured before and after firing. The amount by which it is found to have been shortened measures the pressure to which it has been subjected—a comparison being made as in the Rodman and Woodbridge gauges, with the effect produced by known pressure applied in a testing machine.
The Rodman gauge is used in official experiments in Germany, Russia and Italy. The crusher is used in France and England, and both the crusher and Woodbridge are used in this country.
In ordinary work, chamber pressures only are recorded. For these the gauge is inserted in the face of the mushroom or breech block, or, with a muzzle-loading gun, in a plug fitted to the rear of the powder chamber.
In special experiments, gauges are frequently used in the walls of the gun along the sides of the chamber and the bore.
They are sometimes also screwed into the base of the projectile, but here their record is modified by the motion of the projectile, and to an extent varying with every change in the amount or kind of the powder used.
The following sources of inaccuracy are inherent to all systems relying upon the deformation of metals for the measure of a pressure acting under the conditions which obtain in the bore of a gun:
First There is a wide difference between the effect which a given pressure will produce when applied more or less slowly, as in a testing machine, and that which the same pressure will produce if applied suddenly, as in the explosion of confined gunpowder.
If in the testing machine the force were applied very gradually, and if in the gun the gases were evolved instantaneously, the ratio between the effects produced would be as two to one—the effect in the gun being two. This extreme case is far from existing, both because, by appropriate arrangements, the action of the machine may be made very sudden, and because the development of
the maximum pressure of the powder gas is not instantaneous. Nevertheless, there undoubtedly does exist a difference which, whether large or small, tends to make the indications of the gauge higher than the pressure which has actually existed in the gun.
It should be noted, however, that the pressure is exerted upon the walls of the chamber exactly as it is felt by a pressure gauge in the chamber, so that the real strain upon the gun is that which is shown by the gauge, and not that which would be deduced from a theoretical consideration of the tension of the gases.
Second. In the deformation of a metal, whether by cutting or crushing, the element of time necessarily enters. Up to a certain point the effect will increase with the time.
In a testing machine, the force applied has time to produce its full effect; in a gun the pressure is relieved as suddenly as it is developed. It is not possible to estimate very accurately the time through which the maximum pressure acts, but it probably does not exceed, with the slowest powder, the one thousandth part of a second.
This consideration would lead to the conclusion that the pressures recorded, when interpreted by comparison with the work of a testing machine, are too low. This source of error and the preceding one, then, tend in opposite directions, and to a certain extent offset each other. They would both be reduced to zero if, at the instant when the pressure is at a maximum, it and the force opposed to it by the resistance of the copper discs to deformation were in perfect equilibrium. Such a condition is approached by the device, commonly resorted to in practice, of subjecting the discs which are to be used in experiment to a preliminary pressure slightly lower than that which is anticipated from the firing.
This reduces, but does not remove, the difficulty. Of the error remaining, that part due to the suddenness with which the pressure is applied will be much greater with gauges placed in the path of the projectile forward of the force band than with any others. These are suddenly unmasked by the motion of the projectile to a pressure already developed, which thus comes upon them with the full force of a blow, so that the pressure which they record is doubtless considerably above the tension actually existing. Thus, of two gauges separated from each other by a few inches, but of which one stands in rear of the force band and the other in front of it, the second may give a record several tons higher than that of the first. That this is not a point of theoretical importance only has been abundantly shown in several foreign experiments, notably in those of the English Ordnance Committee of 1875, and more recently in those of Russian officers at Aboukhoff, and of French officers at Gavres. In all of these cases the crusher gauges half a calibre forward of the force band read from twenty per cent, to forty per cent, higher than those in the chamber. With regard to the bearing of these facts upon the strength of guns, it should be considered that, so long as the surface of the bore is continuous, the pressure will be communicated by the elasticity of the metal somewhat in advance of the force band, and no point will receive a sudden blow. At any break, however, which may occur at the continuity of the bore—as, for instance, where the two parts of a tube bust together—there will be an interruption of the transmission of pressure in front of the band. No strain will be felt by that part of the tube forward of the break until the force band has passed it, when the full pressure of the gas will come upon it with a blow, exactly as it would come upon a pressure gauge at that point; and the strain tending to rupture the gun will be just double what it would be if the fibres of the bore were continuous.
This would seem to be a point of weakness in those systems of gun construction in which the tube is composed of two lengths, or in which a liner is inserted which does not extend to the muzzle. It would seem also to furnish an additional reason for the use of a highly elastic steel.
The Deprez manometric balance resembles the instruments already described in opposing a known resistance to the work of the powder gases, but differs from them in the nature of the resistance.
It consists of a piston with a large and a small head connected by a stem running through the wall of the gun. The small head stands flush with the face of the bore, and is acted upon like the piston heads of the gauges already described, by the gases in the gun. The large head is outside the gun and is enclosed by an air chamber, in which it is placed under a known pressure of air. The ratio between the areas of the piston heads is 1 to 400.
Before the gun is fired, the air pressure holds the larger piston head firmly down, binding a small strip of metal between its lower face and a flat plate which forms part of the gauge. A powerful spring attached to the metal strip seeks to draw it out, but can start it only on condition that the piston of the gauge is lifted. If, when the gun is fired, the pressure developed in the bore and exerted upon the small head of the piston exceeds that of the air upon the larger head, the piston will be lifted and the metal strip released. Knowing, then, in any case, the pressure of the air, we can tell of the pressure in the bore, not what it has been, but whether it has or has not exceeded a certain known amount.
By repeating the experiment with constant conditions of powder, projectile, etc., but with varying pressure in the air chamber, it is possible to arrive at a close approximation to the pressure in the bore. It should be added that by an electrical attachment the starting of the metal strip is registered automatically, and that in practice a number of gauges with different pressures are used in a single experiment.
Of instruments which measure pressure by work done in imparting motion, the most important are the Noble chronoscope, the Sebert velocimetre, and the Sebert self-registering projectile.
The Noble chronoscope was used, in connection with crusher gauges, by the English Ordnance Committee which, between 1875 and 1877, conducted a series of experiments upon the action of fired gunpowder. It records the velocity of the projectile at successive intervals of its path in the bore, from which the accelerations, and hence the pressures corresponding, can be deduced. A number of plugs are inserted in the wall of the gun at points forward of the base of the projectile, each plug carrying a small wire which forms part of an electric circuit communicating with a delicate time-recording instrument or chronoscope. When the plugs are in place in the gun, their inner ends, carrying the wires which, as already explained, are in circuit with the recording instrument, stand just inside the bore. The wires are cut in succession by the projectile as it passes through the bore, each circuit, as it is broken, leaving a record upon the cylinder of the chronoscope.
The distances between the plugs being known, and the interval of time between the cutting of any two, the velocity between those two is known, and the pressure acting to produce the successive accelerations can be deduced.
In the Seberf velocimetre the acceleration in velocity of recoil of the gun at a given instant is made the measure of the pressure acting in the bore at that instant.
A ribbon of flexible metal, thinly coated with lampblack, is attached to the gun and moves with it. Above the ribbon stands a tuning-fork, suspended between two electro-magnets. A small spring attached to and vibrating with the tuning-fork makes and breaks the circuits of the electro-magnets so that the latter are magnetized and demagnetized very rapidly, and the tuning-fork is attracted alternately to one side and the other. Thus, after a single instant of initial vibration given to the fork, its motion is controlled by the electromagnets, and will remain constant as long as the strength of the magnets remains constant.
To one of the branches of the tuning-fork is attached a small pencil, held in light contact with the blackened surface of the flexible ribbon. If the fork be started vibrating while the ribbon is at rest, the pencil will trace a line at right angles to the length of the ribbon. As the gun starts to the rear upon firing, the ribbon is drawn out under the tuning-fork and the line traced by the pencil becomes sinuous. From its direction at any given point, knowing the velocity of vibration of the fork, we can determine the velocity of recoil corresponding to that point. From velocities at successive points are deduced accelerations between those points, and from these are derived the pressures acting in .the bore; provided that we have an accurate knowledge not only of the mass of the gun and carriage, but also of the resistance acting in opposition to the recoil.
The Sebert self-registering projectile is a hollow shell, through the centre of which runs a steel rod, square in section, and covered on one of its four sides with a coating of lampblack.
Loosely attached to this rod, and sliding freely upon it, is a block of metal carrying a tuning-fork, to one of the branches of which is fitted a pencil, held lightly in contact with the blackened face of the square rod.
In preparing the projectile for firing, the sliding block is brought to the forward end of the rod, and the branches of the tuning-fork are spread and held apart by a small lug on the rod. Thus prepared, the projectile is placed in the gun like an ordinary shell, and the gun fired. As the projectile starts forward, the sliding block, to which the tuning-fork is fixed, by virtue of its inertia remains at rest, the stem on which it moves sliding through it. At the first instant of motion the lug is withdrawn from between the branches of the tuning-fork, and the latter begins to vibrate at right angles to the direction of the blackened rod, which is slipping past it with the velocity of translation of the projectile. The pencil fixed to the fork thus traces a sinuous line upon the surface of the rod, the direction of which, at any given point, shows the ratio between the velocity of translation of the projectile and the velocity of vibration of the tuning-fork at that point.
It is assumed that friction between the rod and the sliding block is entirely overcome, as practically it is, so that the length of the central rod which has passed the pencil at any given instant is equal to the distance which the projectile has moved up to that instant. From the accelerations indicated by the increase in velocity, the pressures acting on the base of the projectile are deduced—neglecting that part of the work done which is absorbed in friction in the bore and in imparting rotation to the projectile.
As the length of the rod on which the record is made is limited by the length of the projectile, this method can give only the pressures corresponding to the motion of the projectile through something less than its own length. This difficulty has been overcome by a modification of the method in which a second block is set in motion at the forward end of the rod when the first block reaches the base of the projectile.
In the early part of this paper it was stated of the three gauges whose indications depend upon the deformation of metals, that they were subject to sources of error which were reduced, but not entirely removed, by certain precautions commonly taken in practice. It will be interesting, before closing, to inquire how far these errors probably impair the accuracy of results based upon the indications of such gauges; the more so as crusher gauges placed in the powder chamber have been used in the tests of the new Navy guns and powder.
It is sometimes stated that while the indications of such gauges are reliable for comparisons of pressure, they are not to be trusted for absolute values.
This subject has been very elaborately investigated by the eminent French artillerist and mathematician. Monsieur E. Sarrau, who concludes, as the result of his experiments, that the indications of crusher gauges in the powder chamber are practically correct, not only comparatively but absolutely, the errors due to the manner in which the pressure is applied being within the limits of accuracy of the instrument.
With gauges forward of the band slope the case is quite different, and here he concludes that the indications of the gauges are much too light, and altogether unreliable for ballistic calculations.